Many recent evolutionary psychology and human behavioral ecology studies have tested hypotheses by examining correlations between variables measured at a group level (e.g., state, country, continent). In such analyses, variables collected for each aggregation are often taken to be representative of the individuals present within them, and relationships between such variables are presumed to reflect individual-level processes. There are multiple reasons to exercise caution when doing so, including: (1) the ecological fallacy, whereby relationships observed at the aggregate level do not accurately represent individual-level processes; (2) non-independence of data points, which violates assumptions of the inferential techniques used in null hypothesis testing; and (3) cross-cultural non-equivalence of measurement (differences in construct validity between groups). We provide examples of how each of these gives rise to problems in the context of testing evolutionary hypotheses about human behavior, and we offer some suggestions for future research.Keywords Ecological fallacy . Cross-cultural research . Research methods . Simpson's paradox . Non-independence . Measurement equivalence Many hypothesis-testing techniques used by evolutionary behavioral scientists involve examining degrees of covariation between two or more variables and making inferences based on such relationships (e.g., analysis of variance, Pearson correlation). This approach may be applied to a number of different research designs that examine variation at different, hierarchically organized levels (e.g., individuals versus groups;